img
Comparison of T2W FLAIR images of patients with multiple sclerosis and ischemic gliosis via histogram analysis    
Yazarlar
Murat Baykara
Türkiye
Mustafa Yıldırım
Fırat Üniversitesi, Türkiye
Dr. Öğr. Üyesi İsmail TAŞKENT Dr. Öğr. Üyesi İsmail TAŞKENT
Türkiye
Hanefi Yıldırım
Türkiye
Duran Tavut
Özet
Objective: In some cases, it may be challenging to differentiate radiologically between ischemic gliotic foci and multiple sclerosis (MS) plaques. This study aims to evaluate the effectiveness of histogram analysis in the differentiation of MS plaque and ischemic gliosis lesions on the T2-weighted (T2WI) FLAIR sequence. Materials and Methods: This study was conducted on the Magnetic Resonance Imaging (MRI) examinations of patients diagnosed with ischemic gliosis and multiple sclerosis. Inactive lesions of 43 patients with ischemic gliosis and 46 with multiple sclerosis imaged by the same device were included in the study. Histogram analysis parameters of both groups were calculated. The entire image analysis algorithm was obtained through in-house software coded in MATLAB. Both groups were compared using a student’s t-test. The diagnostic value of the parameters was detected with the receiver operating characteristic (ROC) curve. Results: Mean gray level intensity, the standard deviation of the histogram, and entropy values calculated via minimum, maximum and median values were significantly higher in patients with ischemic gliosis. ROC curve analysis indicated that a threshold value of 545.19 for mean gray level intensity has 69.8% specificity and 69.6% sensitivity. Conclusion: Histogram analysis may help differentiate MS and ischemic gliosis.
Anahtar Kelimeler
Multiple Sclerosis | gliosis | image processing | histogram analysis | MRI
Makale Türü Özgün Makale
Makale Alt Türü ESCI dergilerinde yayımlanan tam makale
Dergi Adı Journal of Clinical Practice and Research
Dergi ISSN 2149-2247
Dergi Tarandığı Indeksler ESCI
Makale Dili İngilizce
Basım Tarihi 01-2023
Cilt No 45
Sayı 3
Sayfalar 248 / 252
Doi Numarası 10.14744/etd.2023.35306
Makale Linki http://dx.doi.org/10.14744/etd.2023.35306